Birkett and Phillips Awarded Grant to Help Strengthen HIV Partner Services

Michelle Birkett, Ph.D., and Gregory Phillips II, Ph.D.

After an individual is diagnosed with HIV or other STI, can an interactive digital tool improve the accuracy of identifying who else may have been exposed and the speed of testing and linking those people to care? It’s a question that researchers at Northwestern University’s Institute for Sexual and Gender Minority Health and Wellbeing (ISGMH) want to help answer. ISGMH faculty members Michelle Birkett, Ph.D., and Gregory Phillips II, Ph.D., were recently awarded an R34 grant from the National Institute on Drug Abuse to reconfigure their network data collection software, Network Canvas, to modernize and simplify HIV partner services work at the Chicago Department of Public Health.

Partner services is an evidence-based intervention that connects people recently diagnosed with HIV or STIs to appropriate care and alerts recent sexual or drug use partners who may have been exposed. This partner notification activity has become much more familiar to the general population in the midst of the COVID-19 pandemic—it’s a form of contact tracing that aims to limit the transmission of diseases.

“The primary goal of this HIV/STI contact tracing is understanding who a person’s recent sexual partners and drug partners are so that health departments can contact and offer testing to those partners,” explained Birkett. “We think that contact tracing, particularly within hard-to-reach populations, is where our Network Canvas software can have a real impact on partner services.”

Network Canvas is a free, open-source software suite designed to simplify the collection and use of complex social network data. Created through a collaboration between ISGMH and the University of Oxford, Network Canvas helps researchers capture rich data on individuals and their social environments. Collecting data on an individual’s social and environmental context is particularly useful for understanding what drives the spread of infectious disease. The R34 grant will allow Birkett and Phillips to customize the existing Network Canvas software based on data obtained via a broad national needs assessment and test its viability for use by partner services, first by running a small pilot implementation within the Chicago Department of Public Health.

“There are three different ways we want to change Network Canvas to meet the needs of partner services,” said Birkett. “First, we want to modify existing interfaces so that they are responsive to the needs of both the investigators doing contact tracing as well as to the participants engaged in the activity. Second, we’re thinking about how to deploy this software so that we can reach people as quickly and effectively as possible. This could mean having an investigator use it when interviewing people in person, but it could also mean designing it for interviews done remotely. And the third change is on the backend with our data integration. We want to make sure that what we build links into the existing software that public health departments are already using.”

A number of limitations currently stymie the effectiveness and efficiency of partner services. The methods used by disease investigation specialists—or DIS, the name for people who interview the newly diagnosed and contact their recent partners—are antiquated and lack standardization.

“There is a lot of potential in partner services, but it’s just not currently well suited for 2020,” said Phillips. “Disease investigation specialists may lack proper training, and there is no standardization to how they approach and interview partners who may have been exposed. They’re sometimes called the ‘sex police’ because they’re not always culturally sensitive or culturally appropriate. But at the same time, they’re rarely given the tools needed to do this work effectively.”

DIS are often working with just a pencil, pad of paper, and list of names that have been assigned to them. More modern technologies like text message, email, and online interviews are seldom used, although those methods may be more likely to reach people and encourage participation. The questions asked during partner services interviews tend to differ because DIS don’t have or don’t adhere to identical scripts, even if they work for the same health department. Further, useable data that is collected often remains siloed and is not input into a central, queryable database.

Phillips described the current limitations of disease investigation and how Network Canvas could improve both the collection and utilization of data about sexual and drug partners: “Imagine someone recently diagnosed with HIV telling a disease investigation specialist, ‘I recently had sex with a brown-haired man.’ This isn’t particularly useful for finding that man and connecting him to testing or care. But if a series of standardized questions used by all disease investigation specialists can elicit additional identifying details about that man, we might be able to triangulate information from interviews with other people and find him.”

By using Network Canvas, health departments can both provide their disease investigation specialists with standardized list of questions and, even more critically, gather actionable data that helps them reach people who may need testing and care. Once the initial customization and testing of this software is complete, Birkett and Phillips hope to scale the work up and make the software available to public health departments across the US.